Walia Special Edition on the Bale Mountains (2011) - Zoologische ...

A lot of data we collect in protected areas has a spatial reference. For instance we could count
the number of Ethiopian wolf pups at a den that has a specific location. This is called geographic
data since it describes an event occurring at an explicit position on the earth’s surface. Essentially,
geographical data are any information that can be assigned a unique x and Y coordinate. Geographic
data can be complicated as there are many different descriptions of the earths surface (projections
and datums), however if you are collecting geographic data choose one system and stick with it.
Latitudes and longitudes measured in degrees, minutes, and decimals are the most universal metric
and can be converted to any other projection easily. Most protected areas use this system. If you are
interested in knowing more about geographic data a good overview is compiled in the lecture notes
posted on the Geographer’s Craft website through the University of Colorado (http://www.colorado.
edu/geography/gcraft/contents.html).
Knowing the different types of data is important because as data becomes more detailed it
also becomes increasingly more complicated and expensive to collect (Figure 1). This raises the
question, how do you know at which scale to collect your data?
Which Scale to Collect Data?
Knowing the appropriate level of detail to collect data can be troublesome. Data can be infinitely
precise and as researchers and managers we can be naively drawn to its allure, a bit like a dizzy moth
to a light bulb, but it’s a dangerous game.
To illustrate this point, the Serengeti Ecological Monitoring Program (which is part of the
Serengeti National Park General Management Plan) depends on understanding the relationship
between fires and changes in the vegetation communities. Using long-term data on wildebeest
numbers, fire events and photo points, researchers where able to re-create the complex dynamics
of Acacia forest regeneration (Dublin et al. 1990). The data required were elephant and wildebeest
censuses conducted once every 2 to 5 years for about 40 years, estimates of percent of the area
burnt each year, and a collection of photos taken from the same points every 10 years (roughly).
Now imagine a similar situation where the same data were measured at a finer scale: wildebeest
were counted 10 times a year for 40 years (400 censuses), fires were measured every month for 40
years (480 times), and diameter at breast height for 10,000 trees was measured every year. The
resulting amount of data would be vast, the effort involved in collecting and editing the data would
be immense, and the motivation required in order to analyze such a data set would be heroic. In fact,
chances are that if the data are too complex they will never get used. And if we did analyze it, we
would probably find very similar results to the very simple data collected.
Under the first data collection scheme we would be collecting less detailed data, in a broader
area, on a longer time schedule, and with much less effort. The second scheme, however, would
provide very detailed data, in a very localized geographic area, on a very short time schedule, and
with a huge amount of effort. Therefore, scale can operate both spatially from localized effects to
regional effects, as well as temporally from frequent measures to rare measures. We can think about
<strong>Walia</strong>-<strong>Special</strong> <strong>Edition</strong> on the Bale Mountains 310